We present ClassCast, a tool for class-based forecasting. It partitions an input time series into class-specific time series. It uses conventional prediction methods for each of these time series and maps class-specific values to classified future time indices. It is a simple means to account for non-linear factors in time-series for the purpose of prediction.
CITATION STYLE
Heimgaertner, F., Sachs, T., & Menth, M. (2018). ClassCast: A Tool for Class-Based Forecasting. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 10740 LNCS, pp. 322–326). Springer Verlag. https://doi.org/10.1007/978-3-319-74947-1_26
Mendeley helps you to discover research relevant for your work.